Panorama Tools

Hello
I implemented the Albers Equal Area Conic Projection 
(http://mathworld.wolfram.com/AlbersEqual-AreaConicProjection.html).

Example: http://mdgrosse.net/pano/albersequalareaconic.jpg

Unfortunately I had to change quite a lot in panotool because I needed 
an additional parameter in the p-line of the input script (the phi2) and 
also some persistent information for the transformation functions which 
I both included as array "formatparam" into the Image structure.

If still someone is interested:
http://mdgrosse.net/pano/albersequalareaconic.diff

Example p-line of the input script:
    p f10 P"60.0" w1024 v360  n"TIFF_m"

The P-parameter is the phi2.


Michael

Hi Michael,

Thanks for the implementation. WHen I started implementing new
projections about 3 weeks ago it was clear that some of them needed an
extra parameter.

I look at your diff and it looks complex. Would you mind describing
what you did? perhaps a description in the form of a ChangeLog would
be useful.

Also why did you decide to replace "distance" with "distanceparam"?

thanks!

daniel



 Michael Gross (adv) twisted the bytes to say:

 Michael> Hello
 Michael> I implemented the Albers Equal Area Conic Projection
 Michael> (http://mathworld.wolfram.com/AlbersEqual-AreaConicProjection.html).

 Michael> Example: http://mdgrosse.net/pano/albersequalareaconic.jpg

 Michael> Unfortunately I had to change quite a lot in panotool because I needed
 Michael> an additional parameter in the p-line of the input script (the phi2) and
 Michael> also some persistent information for the transformation functions which
 Michael> I both included as array "formatparam" into the Image structure.

 Michael> If still someone is interested:
 Michael> http://mdgrosse.net/pano/albersequalareaconic.diff

 Michael> Example p-line of the input script:
    p f10 P"60.0" w1024 v360  n"TIFF_m"

 Michael> The P-parameter is the phi2.


 Michael> Michael



 Michael> -------------------------------------------------------------------------
 Michael> Take Surveys. Earn Cash. Influence the Future of IT
 Michael> Join SourceForge.net's Techsay panel and you'll get the chance to share your
 Michael> opinions on IT & business topics through brief surveys - and earn cash
 Michael> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
 Michael> _______________________________________________
 Michael> PanoTools-devel mailing list
 Michael> PanoTools-devel@...
 Michael> https://lists.sourceforge.net/lists/listinfo/panotools-devel

--
Daniel M. German                  "It is useless to punish a man
                                   unless he knows why he is punished...
                                   Punishment must be unusual
   R. Heinlein ->                  or it serves no purpose."
http://turingmachine.org/
http://silvernegative.com/
dmg (at) uvic (dot) ca
replace (at) with @ and (dot) with .

Hello
I updated the diff file. There were some bugs and also it is now 
possible to specify both standard parallels in the p-line, for example:
     p f10 P"60.0 20.0" w1024 v360  n"TIFF_m"

I wrote something about the changes in
   http://mdgrosse.net/pano/albersequalareaconic.txt

I use the MakeParams directly as parameter for the transformation 
functions. With the declared distance-macro it was not possible to use 
mp->distance.

If you have any suggestion, how I could make this parameter-thing better 
(or less "invasive") I will modify it.

Michael

Daniel M. German wrote:
> Hi Michael,
> 
> Thanks for the implementation. WHen I started implementing new
> projections about 3 weeks ago it was clear that some of them needed an
> extra parameter.
> 
> I look at your diff and it looks complex. Would you mind describing
> what you did? perhaps a description in the form of a ChangeLog would
> be useful.
> 
> Also why did you decide to replace "distance" with "distanceparam"?
> 
> thanks!
> 
> daniel
> 
> 
> 
>  Michael Gross (adv) twisted the bytes to say:
> 
>  Michael> Hello
>  Michael> I implemented the Albers Equal Area Conic Projection
>  Michael> (http://mathworld.wolfram.com/AlbersEqual-AreaConicProjection.html).
> 
>  Michael> Example: http://mdgrosse.net/pano/albersequalareaconic.jpg
> 
>  Michael> Unfortunately I had to change quite a lot in panotool because I needed
>  Michael> an additional parameter in the p-line of the input script (the phi2) and
>  Michael> also some persistent information for the transformation functions which
>  Michael> I both included as array "formatparam" into the Image structure.
> 
>  Michael> If still someone is interested:
>  Michael> http://mdgrosse.net/pano/albersequalareaconic.diff
> 
>  Michael> Example p-line of the input script:
>     p f10 P"60.0" w1024 v360  n"TIFF_m"
> 
>  Michael> The P-parameter is the phi2.
> 
> 
>  Michael> Michael
> 
> 
> 
>  Michael> -------------------------------------------------------------------------
>  Michael> Take Surveys. Earn Cash. Influence the Future of IT
>  Michael> Join SourceForge.net's Techsay panel and you'll get the chance to share your
>  Michael> opinions on IT & business topics through brief surveys - and earn cash
>  Michael> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
>  Michael> _______________________________________________
>  Michael> PanoTools-devel mailing list
>  Michael> PanoTools-devel@...
>  Michael> https://lists.sourceforge.net/lists/listinfo/panotools-devel
> 
> --
> Daniel M. German                  "It is useless to punish a man
>                                    unless he knows why he is punished...
>                                    Punishment must be unusual
>    R. Heinlein ->                  or it serves no purpose."
> http://turingmachine.org/
> http://silvernegative.com/
> dmg (at) uvic (dot) ca
> replace (at) with @ and (dot) with .
> 
>  
>

--- "Michael Gross (adv)" <adv@...> wrote:
> I wrote something about the changes in
>    http://mdgrosse.net/pano/albersequalareaconic.txt

thanks! to understand it better, I have a couple of
questions:

P"60.0 20.0"

are these the two standard parallels?
<http://mathworld.wolfram.com/AlbersEqual-AreaConicProjection.html>?

let's call them P1 and P2, as in P"P1 P2".

What is the possible value range for P1 and P2?

does one of them have to be always larger than the
other? or are there no limitations?

is P1=P2 acceptable? if it is not acceptable, will
there be an error code returning?

what are sensible values?

what is the effect of a larger/smaller difference
between P1 and P2?

what is the effect of smaller/larger numbers?

I mean: it is intuitive to me that there will be two
factors changing: the height of the cone (the
development will be more or less vertically stretched)
and its width (which will cover more or less of the
circle). I am trying to understand how the two input
factors relate to these two output observation.

Yuv


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com

Hi
The two parameters are the standard parallels in degrees. The range is=20
from 90=B0 to -90=B0. Along these these parallels there is no angular and=
 no=20
scale distortion. The ordering of the parameters has no effect.

Depending on phi1 and phi2 the shape of the projection is differently=20
curved. You can also make phi1=3Dphi2. When phi1=3Dphi2=3D90 then the=20
projection is circle. With phi1=3Dphi2=3D-90 the projection is also a cir=
cle=20
but the other way round (i.e. 90: tripod is in the center, -90: sky is=20
in the center).

With phi1=3D-phi2 you get also a circle where the zenith and nadir are=20
both located somewhere in the circle. phi1=3D90, phi2=3D0 gets you a=20
half-circle.

A few examples:
   http://mdgrosse.net/pano/albers_-90_-90.jpg
   http://mdgrosse.net/pano/albers_-90_0.jpg
   http://mdgrosse.net/pano/albers_0_-60.jpg
   http://mdgrosse.net/pano/albers_0_10.jpg
   http://mdgrosse.net/pano/albers_0_60.jpg
   http://mdgrosse.net/pano/albers_90_-90.jpg
   http://mdgrosse.net/pano/albers_90_0.jpg
   http://mdgrosse.net/pano/albers_90_90.jpg
   http://mdgrosse.net/pano/albers_9_10.jpg


yuval levy wrote:
> --- "Michael Gross (adv)" <adv@...> wrote:
>> I wrote something about the changes in
>>    http://mdgrosse.net/pano/albersequalareaconic.txt
>=20
> thanks! to understand it better, I have a couple of
> questions:
>=20
> P"60.0 20.0"
>=20
> are these the two standard parallels?
> <http://mathworld.wolfram.com/AlbersEqual-AreaConicProjection.html>?
>=20

Hi Everybody,

I like the projection and  I will incorporate the code (I just need a
few days).

It is important also to remember that some projections are special
cases of others.


For example, oblique projections are a more general projection than
"regular ones", but obliques are not necessary because one can rotate
the panorama before the projection is made (which is not usually done
with the Earth :)

In fact, if we allow the specification of the two parallels, the
ALbert projection is a general case of the Lamberts (both
cylindrical--0,0 I think-- and conic--90,-90).

Pablo, what is your opinion in extending the parser with a
multi-value parameter? I would think that it will be clearer to have 2
parameters instead of 1. This is because some projections need only
one. And this, of course, will affect nona.

dmg



 Michael> Hi
 Michael> The two parameters are the standard parallels in degrees. The r=
ange is=20
 Michael> from 90=B0 to -90=B0. Along these these parallels there is no a=
ngular and no=20
 Michael> scale distortion. The ordering of the parameters has no effect.

 Michael> Depending on phi1 and phi2 the shape of the projection is diffe=
rently=20
 Michael> curved. You can also make phi1=3Dphi2. When phi1=3Dphi2=3D90 th=
en the=20
 Michael> projection is circle. With phi1=3Dphi2=3D-90 the projection is =
also a circle=20
 Michael> but the other way round (i.e. 90: tripod is in the center, -90:=
 sky is=20
 Michael> in the center).

 Michael> With phi1=3D-phi2 you get also a circle where the zenith and na=
dir are=20
 Michael> both located somewhere in the circle. phi1=3D90, phi2=3D0 gets =
you a=20
 Michael> half-circle.

 Michael> A few examples:
   http://mdgrosse.net/pano/albers_-90_-90.jpg
   http://mdgrosse.net/pano/albers_-90_0.jpg
   http://mdgrosse.net/pano/albers_0_-60.jpg
   http://mdgrosse.net/pano/albers_0_10.jpg
   http://mdgrosse.net/pano/albers_0_60.jpg
   http://mdgrosse.net/pano/albers_90_-90.jpg
   http://mdgrosse.net/pano/albers_90_0.jpg
   http://mdgrosse.net/pano/albers_90_90.jpg
   http://mdgrosse.net/pano/albers_9_10.jpg


 Michael> yuval levy wrote:
 >> --- "Michael Gross (adv)" <adv@...> wrote:
 >>> I wrote something about the changes in
 >>> http://mdgrosse.net/pano/albersequalareaconic.txt
 >>=20
 >> thanks! to understand it better, I have a couple of
 >> questions:
 >>=20
 >> P"60.0 20.0"
 >>=20
 >> are these the two standard parallels?
 >> <http://mathworld.wolfram.com/AlbersEqual-AreaConicProjection.html>?
 >>=20


 Michael> ---------------------------------------------------------------=
----------
 Michael> Take Surveys. Earn Cash. Influence the Future of IT
 Michael> Join SourceForge.net's Techsay panel and you'll get the chance =
to share your
 Michael> opinions on IT & business topics through brief surveys - and ea=
rn cash
 Michael> http://www.techsay.com/default.php?page=3Djoin.php&p=3Dsourcefo=
rge&CID=3DDEVDEV
 Michael> _______________________________________________
 Michael> PanoTools-devel mailing list
 Michael> PanoTools-devel@...
 Michael> https://lists.sourceforge.net/lists/listinfo/panotools-devel

--
Daniel M. German                  "If builders built buildings
                                   the way computer programmers
                                   write programs, the first woodpecker
                                   that came along
                                   would have destroyed
    Weinberg's law  ->             all civilization. "
http://turingmachine.org/
http://silvernegative.com/
dmg (at) uvic (dot) ca
replace (at) with @ and (dot) with .

=20

Hi Everybody,

Daniel M. German schrieb:

> For example, oblique projections are a more general projection than
> "regular ones", but obliques are not necessary because one can rotate
> the panorama before the projection is made (which is not usually done
> with the Earth :)
> 
> In fact, if we allow the specification of the two parallels, the
> ALbert projection is a general case of the Lamberts (both
> cylindrical--0,0 I think-- and conic--90,-90).
> 
> Pablo, what is your opinion in extending the parser with a
> multi-value parameter? I would think that it will be clearer to have 2
> parameters instead of 1. This is because some projections need only
> one.

Yes thats probably the best way. For nona, I have already added two letter
parameters, which might also be a good idea for the parameters. So for
example, the new parameters could be named Pa and Pb. If an additional
information is required,

> And this, of course, will affect nona.

For nona it is probably not very complicated. Having a nice interface in
hugin is the bigger question... Maybe two sliders in the preview panel?

Also, for GUI purposes, some way to query the allowed range of these
parameters would be required.

ciao
  Pablo

>> For example, oblique projections are a more general projection than
 >> "regular ones", but obliques are not necessary because one can rotate
 >> the panorama before the projection is made (which is not usually done
 >> with the Earth :)
 >> 
 >> In fact, if we allow the specification of the two parallels, the
 >> ALbert projection is a general case of the Lamberts (both
 >> cylindrical--0,0 I think-- and conic--90,-90).
 >> 
 >> Pablo, what is your opinion in extending the parser with a
 >> multi-value parameter? I would think that it will be clearer to have 2
 >> parameters instead of 1. This is because some projections need only
 >> one.

 Pablo> Yes thats probably the best way. For nona, I have already added two letter
 Pablo> parameters, which might also be a good idea for the parameters. So for
 Pablo> example, the new parameters could be named Pa and Pb. If an additional
 Pablo> information is required,

 >> And this, of course, will affect nona.

 Pablo> For nona it is probably not very complicated. Having a nice interface in
 Pablo> hugin is the bigger question... Maybe two sliders in the preview panel?

 Pablo> Also, for GUI purposes, some way to query the allowed range of these
 Pablo> parameters would be required.

Agreed. I think you need one function to call that returns all the
parameters you need: max vfov, max hfov, number of parms, range of
parms, etc.

Pablo, I think geographers will thank you endlessly for giving them
hugin to play with the way  projections work. Too bad Snyder is
dead. He would have been very excited with this project. I found his
"PRofessional Paper" online:

http://pubs.er.usgs.gov/usgspubs/pp/pp1395

It is in a weird format, so you need to download the plugin. It is a
great treatise in map making.


dmg



  Pablo

 Pablo> -------------------------------------------------------------------------
 Pablo> Take Surveys. Earn Cash. Influence the Future of IT
 Pablo> Join SourceForge.net's Techsay panel and you'll get the chance to share your
 Pablo> opinions on IT & business topics through brief surveys - and earn cash
 Pablo> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
 Pablo> _______________________________________________
 Pablo> PanoTools-devel mailing list
 Pablo> PanoTools-devel@...
 Pablo> https://lists.sourceforge.net/lists/listinfo/panotools-devel

--
Daniel M. German                  "Writing a new OS only for
                                   the 386 in 1991 gets you your
                                   second F for this term. Prof. Andrew S.
                                   Tanenbaum, author of Minix,
                                   in a newsgroup posting
    ->                             to Linus Torvalds."
http://turingmachine.org/
http://silvernegative.com/
dmg (at) uvic (dot) ca
replace (at) with @ and (dot) with .

On Thursday, December 28, 2006 at 2:46, Michael Gross (adv) wrote:

> I implemented the Albers Equal Area Conic Projection
> (http://mathworld.wolfram.com/AlbersEqual-AreaConicProjection.html).
> 
> Example: http://mdgrosse.net/pano/albersequalareaconic.jpg

Nice image, however, I'd be interested how this projection compares 
to the "Orientation plate" remapping described on 
http://wiki.panotools.org/Unusual_remappings
which is already possible with old PTStitcher?

best regards
-- 
Erik Krause
Resources, not only for panorama creation: 
http://www.erik-krause.de/

HI Michael,

I finally got some time to add your code for the Albers
Projection.  Would you mind clarifing the following points?

1. I noticed that you use an unlimited number of "internal parameters",
in this case 7. I suspect you do it for optimization purposes. Is the
idea that some of these parameters can be computed once, instead of in
each iteration? In the current implementation
albersequalareaconic_paramcheck is called every time
albersequalareaconic_erect is called.

2. You seem to interpret North and negative and South as positive
  angles. For instance, then I provide 0 60 (as in your example below)

     http://mdgrosse.net/pano/albers_0_60.jpg

I get the opposite to what I would expect:

http://mathworld.wolfram.com/AlbersEqual-AreaConicProjection.html

To get that I need to use parameters 0, -60

     http://mdgrosse.net/pano/albers_0_-60.jpg

3. What do you think should be the behaviour if only 1 parameter is
   given?=20

Thanks a lot!

--
daniel

    =20
 Michael> Hi
 Michael> The two parameters are the standard parallels in degrees. The r=
ange is=20
 Michael> from 90=B0 to -90=B0. Along these these parallels there is no a=
ngular and no=20
 Michael> scale distortion. The ordering of the parameters has no effect.

 Michael> Depending on phi1 and phi2 the shape of the projection is diffe=
rently=20
 Michael> curved. You can also make phi1=3Dphi2. When phi1=3Dphi2=3D90 th=
en the=20
 Michael> projection is circle. With phi1=3Dphi2=3D-90 the projection is =
also a circle=20
 Michael> but the other way round (i.e. 90: tripod is in the center, -90:=
 sky is=20
 Michael> in the center).

 Michael> With phi1=3D-phi2 you get also a circle where the zenith and na=
dir are=20
 Michael> both located somewhere in the circle. phi1=3D90, phi2=3D0 gets =
you a=20
 Michael> half-circle.

 Michael> A few examples:
   http://mdgrosse.net/pano/albers_-90_-90.jpg
   http://mdgrosse.net/pano/albers_-90_0.jpg
   http://mdgrosse.net/pano/albers_0_-60.jpg
   http://mdgrosse.net/pano/albers_0_10.jpg
   http://mdgrosse.net/pano/albers_0_60.jpg
   http://mdgrosse.net/pano/albers_90_-90.jpg
   http://mdgrosse.net/pano/albers_90_0.jpg
   http://mdgrosse.net/pano/albers_90_90.jpg
   http://mdgrosse.net/pano/albers_9_10.jpg


 Michael> yuval levy wrote:
 >> --- "Michael Gross (adv)" <adv@...> wrote:
 >>> I wrote something about the changes in
 >>> http://mdgrosse.net/pano/albersequalareaconic.txt
 >>=20
 >> thanks! to understand it better, I have a couple of
 >> questions:
 >>=20
 >> P"60.0 20.0"
 >>=20
 >> are these the two standard parallels?
 >> <http://mathworld.wolfram.com/AlbersEqual-AreaConicProjection.html>?
 >>=20


 Michael> ---------------------------------------------------------------=
----------
 Michael> Take Surveys. Earn Cash. Influence the Future of IT
 Michael> Join SourceForge.net's Techsay panel and you'll get the chance =
to share your
 Michael> opinions on IT & business topics through brief surveys - and ea=
rn cash
 Michael> http://www.techsay.com/default.php?page=3Djoin.php&p=3Dsourcefo=
rge&CID=3DDEVDEV
 Michael> _______________________________________________
 Michael> PanoTools-devel mailing list
 Michael> PanoTools-devel@...
 Michael> https://lists.sourceforge.net/lists/listinfo/panotools-devel

--
Daniel M. German                  "You could wind up belivin'=20
                                   that paradise is nothin'
                                    more than a feelin'
   Marillion ->                    that goes on in your mind."
http://turingmachine.org/
http://silvernegative.com/
dmg (at) uvic (dot) ca
replace (at) with @ and (dot) with .

=20

Daniel M. German wrote:
> 1. I noticed that you use an unlimited number of "internal parameters",
> in this case 7. I suspect you do it for optimization purposes. Is the
> idea that some of these parameters can be computed once, instead of in
> each iteration? In the current implementation
> albersequalareaconic_paramcheck is called every time
> albersequalareaconic_erect is called.
Its called every time but in the paramcheck-function there is an if:
- if formatparamcount == 0 (no parameters are given) then they are set 
to default values (0, 60).
- if formatparamcount == 2 (phi1 and phi2) then all the calculation is 
done. This results in 7 parameters.
- if formatparamcount == 7 then nothing is done.
- else an error is returned.


So first time the function is called and the formatparamcount is set to 
7 and the calculation is not done in any subsequent calls.

> 2. You seem to interpret North and negative and South as positive
>   angles. For instance, then I provide 0 60 (as in your example below)
> 
>      http://mdgrosse.net/pano/albers_0_60.jpg
> 
> I get the opposite to what I would expect:
> 
> http://mathworld.wolfram.com/AlbersEqual-AreaConicProjection.html
> 
> To get that I need to use parameters 0, -60
> 
>      http://mdgrosse.net/pano/albers_0_-60.jpg
Yes, I noticed that too and I don't know why this is. By multiplying phi 
and y with -1 you get the same results as in mathworld.


> 3. What do you think should be the behaviour if only 1 parameter is
>    given? 
I would say that then phi1=phi2.

Michael

Erik Krause twisted the bytes to say:

 Erik> On Thursday, December 28, 2006 at 2:46, Michael Gross (adv) wrote:
 >> I implemented the Albers Equal Area Conic Projection
 >> (http://mathworld.wolfram.com/AlbersEqual-AreaConicProjection.html).
 >> 
 >> Example: http://mdgrosse.net/pano/albersequalareaconic.jpg

 Erik> Nice image, however, I'd be interested how this projection compares 
 Erik> to the "Orientation plate" remapping described on 
 Erik> http://wiki.panotools.org/Unusual_remappings
 Erik> which is already possible with old PTStitcher?

I finally have an answer for you :)

They are just different projections. The orientation arch is an
bending of the cylindrical. One could see the Albers Conical Equal
Area as the bending of a cylindrical equal area.

But it will not be possible to generate a ring-like projection with
the albers. At least not without some trick. It will be interesting to
see how it can be used once we can use it within hugin and
interactively play with it.

dmg

 Erik> best regards
 Erik> -- 
 Erik> Erik Krause
 Erik> Resources, not only for panorama creation: 
 Erik> http://www.erik-krause.de/



 Erik> -------------------------------------------------------------------------
 Erik> Take Surveys. Earn Cash. Influence the Future of IT
 Erik> Join SourceForge.net's Techsay panel and you'll get the chance to share your
 Erik> opinions on IT & business topics through brief surveys - and earn cash
 Erik> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
 Erik> _______________________________________________
 Erik> PanoTools-devel mailing list
 Erik> PanoTools-devel@...
 Erik> https://lists.sourceforge.net/lists/listinfo/panotools-devel

--
Daniel M. German                  "Poetry is as exact a science
    Gustave Flaubert ->            as geometry"
http://turingmachine.org/
http://silvernegative.com/
dmg (at) uvic (dot) ca
replace (at) with @ and (dot) with .

I am trying to reduce the size of my mailbox and I finding some older messages.

 yuval> --- "Michael Gross (adv)" <adv@...> wrote:
 >> I wrote something about the changes in
 >> http://mdgrosse.net/pano/albersequalareaconic.txt

 yuval> thanks! to understand it better, I have a couple of
 yuval> questions:

 yuval> P"60.0 20.0"

 yuval> are these the two standard parallels?
 yuval> <http://mathworld.wolfram.com/AlbersEqual-AreaConicProjection.html>?

Yes.

 yuval> let's call them P1 and P2, as in P"P1 P2".

 yuval> What is the possible value range for P1 and P2?

   Presented by H.C. Albers in 1805, takes two parameters (its standard
   parallels, corresponding to the parallels where the cone of projection
   is secant to the sphere). When both are one of the poles then it
   becomes a Lambert Equal Area Azimuthal. When both are the horizon, it
   becomes a Lambert Equal Area Cylindrical. 
   
 yuval> does one of them have to be always larger than the
 yuval> other? or are there no limitations?

Imagine a cone. The cone cuts the sphere at the place where the
parallels are. The order of the parallels does not matter, but their
sign does.

 yuval> is P1=P2 acceptable? if it is not acceptable, will
 yuval> there be an error code returning?

Yes, see above. If P1=P2=90, or 0 then an error is returned. It should
automatically compute the corresponding cylindrical or azimuthal.

 yuval> what are sensible values?

I really don't like the projection :)

 yuval> what is the effect of a larger/smaller difference
 yuval> between P1 and P2?

The cone becomes tangent to the sphere at that point.

 yuval> what is the effect of smaller/larger numbers?

Load an equirectilinear in hugin and play with it.

 yuval> I mean: it is intuitive to me that there will be two
 yuval> factors changing: the height of the cone (the
 yuval> development will be more or less vertically stretched)
 yuval> and its width (which will cover more or less of the
 yuval> circle). I am trying to understand how the two input
 yuval> factors relate to these two output observation.

 yuval> Yuv


 yuval> __________________________________________________
 yuval> Do You Yahoo!?
 yuval> Tired of spam?  Yahoo! Mail has the best spam protection around 
 yuval> http://mail.yahoo.com 

 yuval> -------------------------------------------------------------------------
 yuval> Take Surveys. Earn Cash. Influence the Future of IT
 yuval> Join SourceForge.net's Techsay panel and you'll get the chance to share your
 yuval> opinions on IT & business topics through brief surveys - and earn cash
 yuval> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
 yuval> _______________________________________________
 yuval> PanoTools-devel mailing list
 yuval> PanoTools-devel@...
 yuval> https://lists.sourceforge.net/lists/listinfo/panotools-devel

--
DMG

yuval> is P1=P2 acceptable? if it is not acceptable, will
 yuval> there be an error code returning?

 Daniel> Yes, see above. If P1=P2=90, or 0 then an error is returned. It should
 Daniel> automatically compute the corresponding cylindrical or azimuthal.

After I described it I realized that the Albers becomes a Gall when
P1=-P2.  But the formulae we use for the Albers is undefined when
P1=-P2. We should implement the Gall for the sake of compleness.

dmg

--
Daniel M. German                  "In other studies you go as far as 
                                   others have gone before you,
    ->                             and"
http://turingmachine.org/
http://silvernegative.com/
dmg (at) uvic (dot) ca
replace (at) with @ and (dot) with .